IAES International Journal of Artificial Intelligence (IJ-AI)

Vol. 14, No. 1, February 2025, pp. 73~82

ISSN: 2252-8938, DOI: 10.11591/ijai.v14.i1.pp73-82  73

A page rank-based analytical design of effective search engine

optimization

Vinutha Mysore Srinivas1, Padma Muthalambikasheta Halli Cheluvae Gowda2

1
PET Research Centre, Department of Computer Science and Engineering, University of Mysore, Mandya, India
2
Department of Computer Science and Engineering, PES College of Engineering Mandya, Mandya, India

Article Info ABSTRACT

Article history: Search engine optimization (SEO) is an important internet marketing strategy
and process that facilitates maximizing an intended website’s visibility with
Received Feb 12, 2024 search engine results. It is widely employed nowadays to improve traffic
Revised Jul 2, 2024 volume or quality from search engines to a particular website. Even though a
Accepted Jul 26, 2024 significant number of publications imply the essential aspects of SEO, only a
few provide generalized ideas to deal with the complex structure of the web.
Also, the critical issues of content quality, site popularity, keyword density,
Keywords: and publicity factors were not much considered in the traditional ranking
algorithms during SEO processes. This has negatively influenced the retrieval
PageRank rate in the existing SEO techniques, and consequently, inadequate search
Search engine results were obtained through search engines. Hence, the study considers web
Search engine optimization page ranking as a theoretical basis for the research and addresses these
Web page ranking limitations in the existing system. It further improves SEO performance by
Web pages introducing a unique web-page ranking strategic design to gain higher page
Web structure mining rank results. The results of the investigational study show that the proposed
system effectively contributes towards SEO with an improved page ranking
strategy and also provides higher accuracy in calculating the importance score
of web pages which is comparable with popular ranking algorithms such as
hyperlink-induced topic search (HITS) and PageRank.
This is an open access article under the CC BY-SA license.

Corresponding Author:
Vinutha Mysore Srinivas
PET Research Centre, Department of Computer Science and Engineering, University of Mysore
Mandya, India
Email: [email protected]

1. INTRODUCTION
The concept of a search engine is not new; ‘Archie’ was the first to be released in the early 1990s,
specifically to search file transfer protocol (FTP) data. Conversely, ‘Veronica’ was believed to be the first text-
based search engine ever created [1]. The fundamental motive of businesses in the current digital marketing
era is to communicate appropriate information about their products and services to the right customers through
websites with minimal effort, which has led to an increase in the number of websites on the ‘world wide web’
[2], [3]. Search engine theory develops from the perspective of giving appropriate website pages to the targeted
consumers, and in this way, it is consistent with the fundamental ideas of internet marketing. Search engines
navigate through the billions of pages available on the internet and sort the web pages based on their relevance
to the user-generated query (e.g., keywords and phrases) [4]–[6]. Currently, the popular internet search engines
are Google, Yahoo, and Bing. As a result, whenever a user enters a specific phrase or keyword in a search
instead of the whole website URL for a company, the search engine uses that term to find the appropriate web
pages [7]–[10]. Further, a list with the most pertinent page at the top is displayed. This approach helps
organizations reach their potential consumers by appearing at the top of the search results. Many of the

Journal homepage: http://ijai.iaescore.com

74  ISSN: 2252-8938

traditional search engine optimization (SEO) design strategies are intended to attain higher page rank results
where the popularity of page ranking algorithms arises [11], [12]. Various research studies are being carried
out on designing effective SEO strategies to gain higher page rank results from the perspective of performance
improvement. However, scattered challenges arise when it comes to the practical implementation of research-
based studies in SEO. The research-based studies on SEO have reported various pitfalls associated with the
existing ranking algorithms [13]. Most existing page rank algorithms only employ a few relevant keywords to
retrieve top-k web pages. Also, the resulting web pages may not meet the intended search query. Also, in the
existing page rank algorithms, the need for cost-effective analytical processing with reliable and higher
efficient page ranking is highly envisioned to ensure the adequate performance of SEO [14]–[16].
The research on web structure mining evolves with the purpose of discovering information from the
web. It also searches for information pertaining to relevant scores for web pages and hyperlinks to determine
the quality of the search results. It basically focuses on organizing the hyperlink structure of the web. The study
by Dubey and Roy [17] talks about the significant factors of page ranking for measuring the importance and
behaviour of web pages. Jayaraman et al. [18] also highlights that the performance of SEO could be increased
if the process of website ranking is designed in an optimized flow of execution for particular search terms in
search engines. It also shares the conventional ideas related to SEO on-page ranking while managing the
incoming links and website characteristic features. There is no denial of the fact that the performance of
traditional page rank algorithms could be improved by using web mining techniques, viz., web structure, web
content, and web usage as highlighted in [19].
Alghamdi and Alhaidari [19] also talk about the core web page ranking algorithms and explores their
idea for enhancing the performance of SEO. The idea of page ranking evolved from the most popular baseline
of the PageRank algorithm, which Google employed. The design and operational factors of the traditional
PageRank algorithm utilize the web structure mining concept to compute the page rank values. However, the
core design model of this algorithm computes the rank score of the page at indexing time and evaluates the
page score considering the in-links, which could mislead the search results in the post-ranking phase, as claimed
by Suri et al. [3]. The computational approach of representing the working flow of hypertext-induced topic
selection (HITS) considers a directed graph structure where vertices represent the web pages and a set of edges
depicts the links [20]. Zhang et al. [20] claimed that the performance of this algorithm could be improved to
enhance the scope of SEO operations for various user queries. Chowdhary and Kumar [21], in their study, talk
about the sub-variant of the main PageRank algorithm, which is referred to as the weighted page rank algorithm
(WPR). The authors claim that this baseline approach has a broader scope for improvement owing to its
advantages of computing both in-links and out-links towards measuring the importance of a web page’s score.
Kelotra et al. [22] designed an improved method of page ranking considering the baseline properties
of the traditional PageRank algorithm. Here the method operates on page rank based on the duration a user
spends on the web page and its link structure. The authors claim this approach could effectively offer a better
retrieval rate for web search engines. Hao et al. [23] also enhanced the traditional PageRank algorithm for web
content search. The study by Usha and Nagadeepa [24] also introduces a hybrid page rank algorithm where the
algorithm utilizes web structure, web content, and web usage mining techniques to compute the in-links of the
web pages. Tuteja et al. [25] perform modifications on the design features of traditional WPR considering the
frequency of visits of in-links and out-links, which is further combined with the original mathematical
formulation of the WPR algorithm strategy. Singh and Sharma [26] also introduced another form of page rank
algorithm, which considers both web structure and web usage mining techniques. The algorithm design
computes both in-links and out-links weights and the frequency of visits of in-links on web pages. The study
offers better results with its strategy for improving search engine results. Another approach to page ranking
considering content weight is presented by Joshi and Gupta [27]; here, the study considers content weight
parameters corresponding to web pages for respective query terms to calculate the page ranks. In the study by
Jaganathan and Desikan [28], the page rank algorithm also computes the in-link and out-link weights associated
with the web pages. However, it also constructs its weight matrix, which facilitates retaining the page ranks.
An agent weighted page ranking algorithm (AWPR) is designed as an enhancement of the traditional
WPR algorithm, which is subjected to perform web structure mining while computing the weight of in-links
and out-links presented by Nagappan and Elango [29]. Another similar study by Gupta and Singh [30] also
represents a user preference-based page ranking algorithm where it addresses the topic drift problem of the
AWPR algorithm considering the advantageous factors of web-usage mining in calculating visits of in-link.
Mahajan et al. [31] introduced a extended weighted page rank based on visits of links (EWPR VOL) algorithm
to improve the performance of WPR. Alghamdi and Alhaidari [19] further improvised the performance of
EWPR VOL algorithm to enhance search engine performance for appropriate retrieval of search results
considering web structure, web content and web usage mining approaches. Alhaidari et al. [32] have also
introduced a decision-making trial and evaluation laboratory (DEMATEL) model to improve the performance
of websites towards fulfilling the user’s requirements. A linear programming-based statistical modelling was

Int J Artif Intell, Vol. 14, No. 1, February 2025: 73-82

Int J Artif Intell ISSN: 2252-8938  75

introduced to enhance the ranked list of web search engines by Amin and Emrouznejad [33]. The experimental
outcome shows that optimizing the search engine results takes much longer than usual. Another ranking
technique is designed in the study of Bozkir and Sezer [34], where the approach considers computing visual
similarities among webpages. However, the retrieval scores of a search engine are affected by higher false
positive scores [34]. A similar ranking scheme is also introduced Ahmad et al., [35], where an enumerative
feature, subset–based ranking, modelling was developed to improve the search engine results. However, the
retrieval rate of the scheme was found to be very poor. There are SEO techniques as depicted in the studies in
[36], [37], where the prime emphasis was laid on increasing a website’s visibility. However, the time to mine
the top-ranked website pages was not minimized. Also, the computational complexity is much higher in the
approach of [37]. Ozdemiray and Altingovde [38] introduced a ranking aggregation technique. The primary
motive of this study was to minimize the computational complexity of web page optimizations. However, the
experimental results show that the approach takes comparatively longer to retrieve the top-ranked results by
the search engines. Banaei and Honarvar [14] have encouraged using machine learning-based approaches in
SEO for determining the website’s rank. However, the approach was found satisfactory with the test data but
the retrieval time for extracting top-k web page results was higher due to its iterative operations.
However, the challenge arises to balance the trade-off between computing effort and efficient page
ranking for retrieval of targeted pages on the first page of search results. The analysis of existing approaches
to SEO exhibits that there have been extensive research efforts towards improving the analytical operations in
page ranking strategies. However, it is essential to outline both the strength factors and limitations in the
existing designs of page rank strategies so that readers will have a clearer idea about the scope of improvement
for the future line of research. The analysis of the conventional strategies for SEO-based page ranking
algorithms generates two general research questions (RQs), which are as follows:
‒ RQ1: What contributes to search engine rankings?
‒ RQ2: What can web content creators and admins do to make their content and sites easier to find by
audiences using search engines?
The proposed study also reviewed some essential page ranking strategies in SEO operations.
However, despite having, popularity, the existing page ranking designs in SEO suffers from various
shortcomings, which could be noted as follows:
‒ Most existing page ranking strategies do not cover multiple parameters during SEO operations. This leads
to critical issues regarding content quality, site popularity, keyword density, and publicity factors [17]–[21].
‒ Even though page ranking analytical design modelings are highly encouraged for effective search engine
results towards retrieving top-k targeted pages, the retrieval rate in most of the existing SEO approaches
could be better.
‒ The existing page rank algorithms for SEO need to meet the requirements for intended search as it evaluates
the page score by considering only links. Also, most of the existing SEO designs are affected by the problem
of topic drift [22], [24].
‒ In many cases, challenges arise when dealing with links connected with two or more similar sites. In contrast,
some links could be created unrealistically to encourage the appearance of spam pages by the search engine
(SE) to be in the top-ranked search results. This also misleads the results to the users [29], [30].
‒ It is also observed that the traditional SEO approaches execution workflow models are computationally
complex and affects the retrieval time performance, even though, in many cases, inadequate SE results are
also obtained [35].
‒ The existing SEO approaches need to ensure a proper balance between the retrieval rate of the top-ranked search
results with the computational complexity aspects which influences the time of retrieval [14], [36]–[38].
All the research mentioned above problems are identified to have concrete solutions; hence, the proposed
scheme deploys a novel page rank strategy to address these open-end research problems in SEO.
The study addresses the practical implementation constraints associated with SEO from the cost point
of view and aims to improve the performance of SEO with an optimized page rank strategy. The proposed
study, in this regard, introduces a framework that analytically develops an optimized page rank algorithm to
obtain the significant ranking of web pages so that intended search results could be published with lesser
computational effort. The proposed strategy of page ranking is modeled to facilitate gaining higher page rank
results with an optimized flow of execution and also enhances the retrieval performance while covering
multiple parameters. The uniqueness of the proposed SEO approach lies in the fact that it incorporates novel
analytical strategic execution to contribute towards best search results through improving the performance of
the page rank algorithm and also ensuring adequate convergence value for retrieval of top-ranked targeted
pages within considerable iterations.
The entire manuscript is structured as follows. In section 2, the extensive idea corresponding to the
research methodology is discussed, along with elaborated discussion on the system design, and the algorithm
description. Section 3 highlights the acquired results and justifies the proposed study's effectiveness. Finally,
section 4 also provides conclusive remarks about the overall work and highlights its novel contributions.
A page rank-based analytical design of effective search engine optimization… (Vinutha Mysore Srinivas)
76  ISSN: 2252-8938

2. METHOD
The proposed study introduces a simplified and unique design approach to formulate a page rank
strategy for effective search engine operations. It explores the ideas behind the baseline PageRank models and
further attempts to optimize its flow of execution to normalize ranking a set of websites which also contributes
towards effective search engine results. Here the prime motive of this algorithm is to make the targeted search
results appear on the top of the front page with the pages having the highest relevancy scores. The conventional
SEO designs are based on the PageRank algorithm variants, which aim to retain high page rank scores at the
top of the front page considering a specific user’s provided search (query).
The study applies a power method that updates the weighted reference counts generated by the
hyperlinks between pages. Further, the system applies a connectivity vector and evaluates the hyperlinks
between the pages to measure the in-degree and out-degree for the respective pages. The computation further
also explores the probabilistic factors of user behaviour and constructs a transitional probabilistic matrix.
Further, the strategy applies Perron–Frobenius theorem and assesses the scaling factor to obtain the page rank
score for respective pages. The study also applies a normalization technique to make the ranking of the web
pages more reliable within considerable amount of iteration for execution schema. The page scores are further
normalized considering the total count of out-going links of the source nodes. The strategy also shares the idea
of normalizing page rank of each page considering a mean value operation and further assesses the iterations
to retain highest page rank scores for more significant pages during SEO operations.
The proposed approach’s prime motive is to design and develop a novel analytical operation for page
ranking in SEO. An explicit system design and modeling approach is constructed to realize the system
operations for effective search engine results from both retrieval performance and cost of computation point of
view. The system design modeling considers graph theory to explore the link structure of world wide web
(www). Here the system design and modeling are also inspired from the conventional PageRank algorithm
towards ranking the search engine results. The study also explores the idea through which PageRank algorithm
ranks a collection of websites.

2.1. Modelling of markov process

The study considers the theory of the Markov chain or Markov process [39] to design the analytical
operations of page ranking in SEO. The underlying idea of Markov chain refers to a stochastic process in which
states change for the transition probabilities. Here the transition probabilities are determined by the steps of the
previous time step. This theory is crucial to model large-scale systems with random behaviour where the area
of search engine operation also arises. In web surfing, a user navigates from one page to another by randomly
choosing the outgoing links. This can lead to the dead end of web pages with no outgoing links. Alternatively,
it can also happen that the user cycles around interconnected pages. So, it is evident that a user tends to choose
a random page from the web in a certain fraction of time. This scenario is often called a random walk and
theoretically can be described with Markov process.

2.2. Analytical operation of pagerank strategy

A probabilistic strategic evaluation is related to estimating page rank score. Here, the proposed system
considers a limited probabilistic score (𝑃𝑙 ), which denotes the likelihood that a random web surfer will visit
any website. It also goes by the name PageRank. Let W is a set of web pages which can be represented as
𝑊 = {𝑊𝑖 }𝑛𝑖=1 . Here 𝑛 represents the number of web pages. The formulated page ranking strategy basically
operates on the count of incoming (𝐼𝑙𝑖𝑛𝑘 ) and outgoing links (𝑂𝑙𝑖𝑛𝑘 ) to a page. The strategy also evaluates the
quality of links to a page for effective SEO. This idea helps determining and generalizing the degree of
importance (DOI) of a particular website. For example, a graph-based representation can be modeled to depict
three web pages such as 𝑊 = {𝑊𝑖 }3𝑖=1 in the form of three vertices. The webpages can be reached through the
formulation of hyperlinks (ℎ𝑙 ) which begins at any of the root page. The proposed ranking strategy initially
constructs the connectivity vector and further enables an explicit functional module ƒ1 (𝑥) to compute the row
and column summation from the 𝑉𝑛×𝑛 . The system also computes a perceived importance factor (𝐼𝐹 ) for a
particular website or web-page through the proposed page rank strategy for effective SEO. The proposed page
rank strategy is implemented over SEO for effective search engine results considering graph-based modeling.
The connectivity vector 𝑉𝑖𝑗 represents connection or hyperlink between page 𝑖 to page 𝑗. Also, the strategy
further estimates a probability factor of 𝑃𝑟 , which implies the possibility of an Internet user to randomly select
and follows a link of a current page. Another probability measure of 𝛿 implies the possibility of choosing a
specific random page which can be computed as (1):

(1 − Pr )⁄
δ= n (1)

Int J Artif Intell, Vol. 14, No. 1, February 2025: 73-82

Int J Artif Intell ISSN: 2252-8938  77

Here the 𝑉𝑛×𝑛 represents a connectivity matrix which corresponds to a portion of web structure. The
proposed formulation of page rank strategy also estimates the quantities of 𝑟𝑜𝑤𝑗 and 𝑐𝑜𝑙𝑗 which indicates the
in-degree and out-degree measure of 𝑗𝑡ℎ page. The strategic solution further also constructs a matrix 𝑇 which
is also of the dimension (𝑛 × 𝑛). The elements of 𝑇𝑖𝑗 can be represented with the (2).

Pr Vij
+ δ ∶ colj ≠ 0
colj
Tij = { (2)
1
∶ colj = 0
n

Here the computation of 𝑇𝑖𝑗 takes place by scaling the connectivity matrix 𝑉𝑖𝑗 with respect to its
column sums. Here the 𝑗𝑡ℎ column in 𝑇𝑖𝑗 indicates the possibility of an user jumping from one page to another
pages in the web. If it is found that the 𝑗𝑡ℎ page is dead end then it has not out-going links to be associated.
The strategy applies a uniform probability factor of 1/𝑛 in all the elements of the column vector. It can be seen
that most of the elements in the matrix 𝑇𝑖𝑗 belongs to 𝛿 that indicates the possibility of jumping from one page
to another without following a link. Here the transition probability matrix 𝑇𝑖𝑗 is computed considering the
theory of Markov chain. The characteristics of this 𝑇𝑖𝑗 is that their element lies between {0,1} and its columns
sum is computed as 1. The study further also applies another explicit functional modeling ƒ2 (𝑥) to compute
𝑇𝑖𝑗 . The proposed study further also employs a methodology of Perron–Frobenius theorem [40] to the matrix
which is retained. The study here applies another functional module f3(x) to compute the non-zero solution.
The study also explores the best way to optimize the computation of page rank strategy by exploring advantage
of particular structure of Markov matrix. Here the approach attempts to preserve the sparsity factor associated
with 𝑉𝑖𝑗 . The formulation of the computation of transition matrix can be formed as (3):

T = ∑(P ∗ V ∗ Diag, ε) (3)

Here 𝐷𝑖𝑎𝑔 represents a diagonal matrix(𝑑𝑗𝑗 ) considering the reciprocals of the out-degrees. Also 𝜀
implies a rank one matrix which is accounted for the random choices of web pages that do not follow the links.
Here the page rank strategy can be optimized with the (4).

(I − P ∗ V ∗ Diag)x = κ (4)

Here 𝜅 is the 𝑛 vector of all ones and correspond to 𝜀. The progressive computation of page ranking
can be further updated with respect to the following normalized expression:

(1−Pr ) r. s
r=∑ , Pr × (M ′ × ( ) + ) (5)
n d n

Here 𝑟 denotes a vector consisting of page rank scores whereas 𝑃𝑟 implies a scalar dumping factor and
its value is considered to be 0.85. This probability factor indicates the possibility of a user to click on a link on
a current page rather continuing to another random page. Here 𝑀′ represents an adjacency matrix of the web
graph structure. Also, the vector d indicates the out-degree measure of a node in the graph structure. The value
of d is considered to be 1, if there exist nodes with no outgoing links. 𝑛 represents the scalar number of nodes
in the graph. Here in expression (5), 𝑠 represents sum of the page rank scores for the pages having no links.

2.3. Algorithm design for page ranking in search engine optimization

The following analytical algorithm exhibits the workflow model of the proposed algorithm design
strategy of page ranking for SEO. Here study formulates simplified work flow modeling of the design of page
ranking strategy for effective SEO. The study also incorporates a set of explicit functionalies to model the
design strategy of SEO where a set of baseline strategies are also reffered for optimized execution. The steps
associated with the proposed ranking strategy are illustrated in Algorithm 1.

Algorithm 1: For page ranking for effective SEO

Input: source(s), target (t)
Output: rank r
Start
1. Init s,t
2. {Wi }ni=1 for s,t
3. For i=1:s

A page rank-based analytical design of effective search engine optimization… (Vinutha Mysore Srinivas)
78  ISSN: 2252-8938

4. For j = 1:t
5. Vn×n , Ilink , Olink , digraph
6. Execute ƒ1 (𝑥)
7. rowi , colj Pr ,1 − Pr , δ
8. ƒ2 (𝑥) to compute Tij (Markov Chain)
9. Perron–Frobenius theorem [40]
10. Optimize transition matrix T
11. (I − P ∗ V ∗ Diag)x = κ (5)
12. Compute Page Rank r for SEO using (5)
13. Normalization of each r
14. Retain Significant Page Rank Score
15. End
End

The above analytical operations involved in the proposed page ranking strategy are applied over SEO
for effective SE results for target and relevant top page retrieval. The study also applies formulation of directed
graph structure to visualize the web structure model to illustrate how each node representing web page confers
its specific rank score to other nodes or web pages. Unlike existing page rank strategies (HITS and PageRank),
the proposed idea of the simplified and light-weight analytical framework of SEO considering the optimized
page rank strategy not only enhance the retrieval performance, also the reduced iterations for optimal
processing of the work flow model ensures effective retrieval time.
The novelty of the proposed page rank algorithm of SEO is as follows:
‒ Unlike existing page rank strategies, the proposed page rank algorithm contributes towards enhancing the
retrieval efficiency of SE for target top-k pages.
‒ The design idea is simplified for light-weight analytical operations which also ensures cost effective
computation and shorter retrieval time for SEO
‒ Unlike existing system, the proposed algorithm handles the problem of topic drift and set high rank values
to more popular pages for effective search engine results on the top of the first page.
A closer look into the entire algorithm implementation shows that the proposed scheme offers a novel and
sophisticated SEO operations with balanced performance between retrieval efficiency and retrieval time. The
next section further illustrates the experimental outcome obtained from a strategic implementation of the
formulated optimized page rank algorithm of SEO.

3. RESULT ANALYSIS
This section illustrates the numerical outcome obtained after simulating the proposed algorithm over
six different websites 𝑊. The prime reason behind adopting analytical strategy for numerical simulation is – it
provides better representation of the outcome considering different metrics through which the effectiveness of
the proposed strategy could be validated to a greater extent. The result analysis also covers the simulation
assessment strategy along with experimental outcome and analysis to conclude the effectiveness of the study
model.

3.1. Simulation assesment strategy

The study considers MATLAB to construct the framework modeling for proposed optimized page
ranking of SEO. It considers a regular 64-bit Windows machine with i5 processing capability. The algorithm
is strategically modeled and scripted considering analytical schema to realise the objective of the proposed
research study. The numerical analysis is considered to compute the values and to visualize the outcome as
obtained from the proposed algorithm. For the purpose of performance assessment of the proposed strategic
schema of page ranking the study not only relies on evaluating the page rank score for web pages but it also
considers computational time as a parameter of complexity in the form of number of iterations of the algorithm
to judge the how it converges towards the targeted top-retrieval of pages in SE results in considerable amount
of retrieval time. The experimental outcome is further assessed for a comparative study under different
conditions.

3.2. Experimental assesment and analysis

The strategy for result analysis considers implementing this proposed page rank idea to enhance the
search engine performance for both retrieval efficiency with respect to ranking and reducing the computational
time to minimize the time of retrieval. The Figure 1 shows the page rank measure obtained through the

Int J Artif Intell, Vol. 14, No. 1, February 2025: 73-82

Int J Artif Intell ISSN: 2252-8938  79

proposed page rank algorithm of SEO towards retrieving the top-8 pages from the SE. The degree of
information of node computes the average InDegree and OutDegree measure are 20 and 14 respectively for the
retrieved pages. The proposed study also further normalizes the page rank score of the individual pages to
enhance the SE results considering the (6).

r(i)
r̅i = (6)
μ∀r

Figure 1. PageRank measure for top-k pages retreived in SE

The normalization of page rank score significantly improves the performance of SE. The study further
analyses the graph structure of connectivity for degree of information among the nodes. The study also
extended the analysis of the proposed optimized page rank algorithm of SEO for implementing it on another
website having set of webpages linked. In this regard it refers to a data which is obtained through automatic
web crawler for a particular website. The experiment here constructs an adjacency matrix of Vij where the
connections between 100 nodes are explored and their information are extracted. The nodes in the clique are
highly connected with each other and it also depicts the fact that if a random web user uses SE, then that user
has got a probability factor of approximately 4.5% chance to arrive on the retrieved pages. The study also
evaluates the measure of computational time complexity for iterations which also reflects the algorithm’s
influence on minimizing the retrieval time.
The Figure 2 shows the comparative analysis among the conventional page rank algorithms and the
proposed PageRank algorithm for the measure of computational time complexity measure in average
processing time (sec). The Figure 2 shows that the proposed optimized PageRank strategy converges towards
the target retrieved pages with minimal number of iterations which is comparatively more in the case of
conventional PageRank algorithm and HITS. One of the primary reasons of optimizing the iteration process is
that the proposed PageRank strategy simplifies the execution process with explicit functions where the core
computational efforts are reduced to a greater extent. It also applies a normalization function for page ranking
and retrieve the most significant pages in SE. This is how the proposed strategy also reduces the retrieval time
for SE results. The study also performs a comparative analysis with significant related studies as highlighted
in the literature section of the study is shown in Table 1. Table 1 highlights the comparison with the state-of-
the-art methods presented in section 1. From the outcome shown in Table 1, the following inference of novelty
is drawn:
‒ The closer analysis of the performance of the proposed PageRank strategy in SEO shows that unlike
conventional approaches it accomplishes a well-balanced performance between Retrieval Accuracy and
Retrieval Time. The performance of Retrieval Time is influenced by optimizing the iterations of the
proposed PageRank algorithm design.
‒ The study also shows that the simplified design modeling of PageRank strategy is cost-effective and has
reduced dependencies of iterative computation towards reaching to the convergence factor of the high-
ranked target page on SE.
A page rank-based analytical design of effective search engine optimization… (Vinutha Mysore Srinivas)
80  ISSN: 2252-8938

‒ Unlike conventional machine learning based designs for SEO, the proposed PageRank strategy also
optimizes the retrieval time with simplified and progressive analytical operations which makes it more
significant and lesser complex.
‒ The light-weight analytical operations not only ensure better retrieval score of PageRank but also reduces
the complexity execution modeling as comparison to the existing PageRank strategy.

Figure 2. Comparative study of computing time measure for iterations

Table 1. Comparison with state-of-art

Method Computational complexity Retrieval accuracy Retrieval time
ProP (PageRank) Low High Low
Alghamdi and Alhaidari [19] High High High
Suri et al. [20] High Low High
Jayaraman et al. [18] Moderate Low Moderate
Banaei and Honarvar [14] High Moderate High
Alhaidari et al. [32] High Moderate High

4. CONCLUSION
The study in this research work introduces an optimized PageRank algorithm for effective search
engine operations. It designs the SEO based on a normalized PageRank analytical strategy where the flow of
execution is optimized to balance the performance of retrieval efficiency for top-k pages and the retrieval time.
This research approach offers an extensive study of the conventional related work on page ranking and further
derives the problem from its core findings. The proposed research methodology uses the advantageous factors
of Markov process modeling and the Perron–Frobenius theorem to offer better search results over SE. Unlike
the existing system, the study also offers a well-balanced performance between retrieval efficiency and retrieval
time by normalizing the page rank scores, significantly reducing the convergence time for retaining the top-k
pages through SEO. In the end, the study also compares the performance of the proposed PageRank strategy
with the state-of-the-art page rank designs. It justifies the proposed PageRank strategy's effectiveness in
enhancing SEO performance. Future research will explore the formulated PageRank algorithm with more
varying execution parameters and analyze its performance.

REFERENCES
[1] L. Page and S. Brin, “The anatomy of a large-scale hypertextual Web search engine,” Computer Networks, vol. 30, no. 1–7, pp.
107–117, 1998, doi: 10.1016/s0169-7552(98)00110-x.
[2] R. S. Bhandari and A. Bansal, “Impact of search engine optimization as a marketing tool,” Jindal Journal of Business Research,
vol. 7, no. 1, pp. 23–36, 2018, doi: 10.1177/2278682117754016.
[3] S. Suri, A. Gupta, and K. Sharma, “Comparative analysis of ranking algorithms used on web,” Annals of Emerging Technologies
in Computing, vol. 4, no. 2, pp. 14–25, 2020, doi: 10.33166/AETiC.2020.02.002.

Int J Artif Intell, Vol. 14, No. 1, February 2025: 73-82

Int J Artif Intell ISSN: 2252-8938  81

[4] N. Papagiannis, Effective SEO and content marketing: the ultimate guide for maximizing free web traffic. Hoboken, USA: John
Wiley & Sons, 2020.
[5] S. Gupta, N. Agrawal, and S. Gupta, “A review on search engine optimization: basics,” International Journal of Hybrid Information
Technology, vol. 9, no. 5, pp. 381–390, 2016, doi: 10.14257/ijhit.2016.9.5.32.
[6] M. R. Baye, B. D. L. Santos, and M. R. Wildenbeest, “Search engine optimization: what drives organic traffic to retail sites?,”
Journal of Economics and Management Strategy, vol. 25, no. 1, pp. 6–31, 2016, doi: 10.1111/jems.12141.
[7] A. Veglis and D. Giomelakis, “Search engine optimization,” Future Internet, vol. 12, no. 1, 2020, doi: 10.3390/fi12010006.
[8] T. Kelsey, Introduction to search engine optimization-a guide for absolute beginners. Berkeley, California: Apress, doi:
10.1007/978-1-4842-2851-7.
[9] P. Rani and E. S. Singh, “An offline SEO (Search engine optimization) based algorithm to calculate web page rank according to
different parameters,” International Journal of Computers & Technology, vol. 9, no. 1, pp. 926–931, 2013, doi:
10.24297/ijct.v9i1.4161.
[10] A. Jain, “The role and importance of search engine and search engine optimization,” International Journal of Emerging Trends &
Technology in Computer Science (ITETTCS), vol. 2, no. 3, pp. 1–102, 2013.
[11] M. N. A. Khan and A. Mahmood, “A distinctive approach to obtain higher page rank through search engine optimization,” Sadhana
- Academy Proceedings in Engineering Sciences, vol. 43, no. 3, 2018, doi: 10.1007/s12046-018-0812-3.
[12] E. D. Ochoa, An analysis of the application of selected search engine optimization (SEO) techniques and their effectiveness on
Google’s search ranking algorithm, Northridge: California State University, 2012.
[13] N. Craswell, F. Crimmins, D. Hawking, and A. Moffat, “Performance and cost tradeoffs in web search,” in Proceedings of the 15th
Australasian database conference, 2004, pp. 161–169.
[14] H. Banaei and A. R. Honarvar, “Web page rank estimation in search engine based on SEO parameters using machine learning
techniques,” International Journal of Computer Science and Network Security, vol. 17, no. 5, pp. 95–100, 2017.
[15] S. Zhang and N. Cabage, “Search engine optimization: Comparison of link building and social sharing,” Journal of Computer
Information Systems, vol. 57, no. 2, pp. 148–159, 2017, doi: 10.1080/08874417.2016.1183447.
[16] J. Zilincan, “Search engine optimization,” in CBU International Conference Proceedings, vol. 3, 2015, pp. 506–510, doi:
10.12955/cbup.v3.645.
[17] H. Dubey and P. B. N. Roy, “An improved page rank algorithm based on optimized normalization technique,” International journal
of Computer Science and Information technology (IJCSIT), vol. 2, no. 5, pp. 2183–2188, 2011.
[18] S. Jayaraman, M. Ramachandran, R. Patan, M. Daneshmand, and A. H. Gandomi, “Fuzzy deep neural learning based on goodman
and Kruskal’s gamma for search engine optimization,” IEEE Transactions on Big Data, vol. 8, no. 1, pp. 268–277, 2022, doi:
10.1109/TBDATA.2020.2963982.
[19] H. Alghamdi and F. Alhaidari, “Extended user preference based weighted page ranking algorithm,” 2021 IEEE 4th National
Computing Colleges Conference, NCCC 2021, 2021, doi: 10.1109/NCCC49330.2021.9428844.
[20] X. Zhang, H. Yu, C. Zhang, and X. Liu, “An improved weighted HITS algorithm based on similarity and popularity,” 2nd
International Multi-Symposiums on Computer and Computational Sciences, IMSCCS’07, pp. 477–480, 2007, doi:
10.1109/IMSCCS.2007.4392647.
[21] A. Chowdhary and A. Kumar, “Study of web page ranking algorithms: a review,” Acta Informatica Malaysia, vol. 3, no. 2, pp. 1–
4, 2019, doi: 10.26480/aim.02.2019.01.04.
[22] A. Kelotra, “Upgradation of PageRank algorithm based upon time spent on web page and its link structure,” International Journal
of Computer Applications, vol. 109, no. 11, pp. 7–9, 2015, doi: 10.5120/19230-0952.
[23] Z. Hao, P. Qiumei, Z. Hong, and S. Zhihao, “An improved pagerank algorithm based on web content,” 14th International
Symposium on Distributed Computing and Applications for Business, Engineering and Science, DCABES 2015, pp. 284–287, 2016,
doi: 10.1109/DCABES.2015.78.
[24] M. Usha and N. Nagadeepa, “A hybrid web page ranking algorithm to achieve effective organic search result,” International Journal
of Scientific Research in Science and Technology, vol. 3, no. 6, pp. 392–402, 2017.
[25] S. Tuteja, “Enhancement in weighted PageRank algorithm using VOL,” IOSR Journal of Computer Engineering, vol. 14, no. 5, pp.
135–141, 2013, doi: 10.9790/0661-145135141.
[26] R. Singh and D. K. Sharma, “Enhanced-ratioRank: enhancing impact of inlinks and outlinks,” 2013 IEEE Conference on
Information and Communication Technologies, ICT 2013, pp. 287–291, 2013, doi: 10.1109/CICT.2013.6558107.
[27] R. Joshi and V. K. Gupta, “Improving PageRank calculation by using content weight,” International Journal of Science and
Research (IJSR), vol. 3, no. 6, pp. 2319–7064, 2014.
[28] B. Jaganathan and K. Desikan, “Weighted page rank algorithm based on in-out weight of webpages,” Indian Journal of Science and
Technology, vol. 8, no. 34, 2015, doi: 10.17485/ijst/2015/v8i34/86120.
[29] V. K. Nagappan and P. Elango, “Agent based weighted page ranking algorithm for web content information retrieval,” Proceedings
of the International Conference on Computing and Communications Technologies, ICCCT 2015, pp. 31–36, 2015, doi:
10.1109/ICCCT2.2015.7292715.
[30] D. Gupta and D. Singh, “User preference based page ranking algorithm,” IEEE International Conference on Computing,
Communication and Automation, ICCCA 2016, pp. 166–171, 2017, doi: 10.1109/CCAA.2016.7813711.
[31] I. Mahajan, S. Gupta, H. Kaur, and D. Kumar, “Extended weighted page rank based on VOL by finding user activities time and
page reading time, storing them directly on search engine database server,” International Journal of Engineering Works, vol. 4, no.
2, pp. 41–48, 2017.
[32] F. Alhaidari, S. Alwarthan and A. Alamoudi, “User preference based weighted page ranking algorithm,” 2020 3rd International
Conference on Computer Applications & Information Security (ICCAIS), Riyadh, Saudi Arabia, 2020, pp. 1-6, doi:
10.1109/ICCAIS48893.2020.9096823.
[33] G. R. Amin and A. Emrouznejad, “Optimizing search engines results using linear programming,” Expert Systems with Applications,
vol. 38, no. 9, pp. 11534–11537, 2011, doi: 10.1016/j.eswa.2011.03.030.
[34] A. S. Bozkir and E. A. Sezer, “Layout-based computation of web page similarity ranks,” International Journal of Human Computer
Studies, vol. 110, pp. 95–114, 2018, doi: 10.1016/j.ijhcs.2017.10.008.
[35] M. W. Ahmad, M. N. Doja, and T. Ahmad, “Enumerative feature subset based ranking system for learning to rank in presence of
implicit user feedback,” Journal of King Saud University - Computer and Information Sciences, vol. 32, no. 8, pp. 965–976, 2020,
doi: 10.1016/j.jksuci.2017.12.006.
[36] J. B. Killoran, “How to use search engine optimization techniques to increase website visibility,” IEEE Transactions on Professional
Communication, vol. 56, no. 1, pp. 50–66, 2013, doi: 10.1109/TPC.2012.2237255.
[37] X. Gui, J. Liu, Q. Lv, C. Dong, and Z. Lei, “Probabilistic top-k query: Model and application on web traffic analysis,” China
Communications, vol. 13, no. 6, pp. 123–137, 2016, doi: 10.1109/CC.2016.7513208.
A page rank-based analytical design of effective search engine optimization… (Vinutha Mysore Srinivas)
82  ISSN: 2252-8938

[38] A. M. Ozdemiray and I. S. Altingovde, “Explicit search result diversification using score and rank aggregation methods,” Journal
of the Association for Information Science and Technology, vol. 66, no. 6, pp. 1212–1228, 2015, doi: 10.1002/asi.23259.
[39] H. Ishii and R. Tempo, “Markov chains and ranking problems in web search,” Encyclopedia of Systems and Control, pp. 691–697,
2015, doi: 10.1007/978-1-4471-5058-9_135.
[40] S. U. Pillai, T. Suel, and Seunghun Cha, “The Perron-Frobenius theorem: some of its applications,” IEEE Signal Processing
Magazine, vol. 22, no. 2, pp. 62–75, 2005, doi: 10.1109/msp.2005.1406483.

BIOGRAPHIES OF AUTHORS

Vinutha Mysore Srinivas received B.E. degree in Electrical & Electronics

Engineering from VTU, M.E. in Information Technology from Bangalore university. She is
currently pursuing the Ph.D. in the Department of Computer science & Engineering, PET
Research foundation, PESCE, Mandya Affiliated to University of Mysore. She is currently
Assistant Professor in Department of Computer Science & Engineering at Dr Ambedkar Institute
of Technology, Bengaluru, Karnataka, India. She has published several papers in various reputed
international journals and conferences. Her current research interest includes data science,
natural language processing, data analytics, and graph theory. She can be contacted at email:
[email protected].

Dr. Padma Muthalambikasheta Halli Cheluvae Gowda received the B.E. degree
in computer science and engineering from PESCE, Mandya, M.Sc. Tech from University of
Mysore and Ph.D. from Visvesvaraya Technological University (VTU), Belgaum. She is
currently professor in Department of Computer science & Engineering and Dean Research,
PESCE, Mandya. She has served on several panels of academic bodies for universities and
autonomous colleges as a BOS and BOE member. She has published several papers in various
reputed international journals and conferences. Her research interests include in pattern
recognition, image processing (document image analysis and recognition), natural language
processing, data mining, data analytics, and sentiment analysis. She is serving as an editor and a
reviewer for various prestigious international journals. She has delivered many keynote
addresses and invited talks throughout India on a variety of topics related to computer science
and engineering. She can be contacted at email: [email protected].

Int J Artif Intell, Vol. 14, No. 1, February 2025: 73-82